The present invention relates to network management, and more specifically, to diagnostic systems in distributed networks.
Network management typically involves an administrator analyzing network points for debugging and diagnosis of networks performance issues. For example, an administrator may check if traffic is balanced over servers in a switch or if latency is unusually high along a data path. For the administrator to diagnose and debug a performance issue, he or she typically has to log in to the switch under inspection and review data logs. If multiple switches on a network are being affected, then the administrator may need to log in to each switch and review the logs of each. The logs may show performance on a component by component basis. Additional information may include Boolean checks, for example, whether network components are operating within threshold criteria. In a distributed or scaled out network, this can mean physically inspecting several switching points for example, 128 switch boxes of multiple switches, and analyzing the data logs for diagnosis and debugging. Furthermore, the administrator may have to perform diagnosis based on multiple disparate parameters and issues.